Effect of maternal position and uterine activity on periodic maternal heart rate changes before elective cesarean section at term
Abstract
Introduction
Because little is known about the effects of maternal position on periodic changes in the maternal heart rate (MHR) in late pregnancy, a prospective observational study was done at Tygerberg Academic Hospital in Cape Town. Pregnant women admitted for elective cesarean section were studied to determine the effect of changes in position on the maternal and fetal heart rates (FHR) and maternal blood pressure.
Material and methods
Continuous transabdominal non-invasive recording of MHR, FHR patterns and uterine activity was done for 1 h in 119 women, using the AN24 device from Monica Health Care. Maternal position was changed every 15 min from lateral to supine, then to the other lateral position and finally supine again. Blood pressure was measured in the left arm and left lower leg three times during each 15-min period.
Results
MHRs were four beats per minute slower in the left lateral position than in the right lateral position. Periodic MHR changes were seen in 13 (10.9%) women. Most of these (84.6%) were associated with uterine activity and not with maternal position. No changes in FHR patterns were observed after position changes.
Conclusions
In a subgroup of pregnant women at term, uterine activity was associated with periodic decelerations of the MHR. In low risk pregnancies there seems to be no effect on the FHR pattern. Implications for the compromised fetus have not yet been investigated.
Abbreviations
-
- bpm
-
- beats per minute
-
- FHR
-
- fetal heart rate
-
- MHR
-
- maternal heart rate
Key Message
Maternal heart rate (MHR), fetal heart rate and uterine activity were recorded by transabdominal fetal electrocardiogram in 119 pregnant women before elective cesarean section. Regular periodic changes in the MHR were observed in 10.9% of these women. Most of these changes (84.6%) were associated with uterine activity.
Introduction
Periodic changes in the maternal heart rate (MHR) have previously been described in the setting of labor. A literature review identified no published studies pertaining to the effect of uterine activity on the MHR prior to the onset of labor.
Coincidental monitoring of the MHR was first described in cases of fetal demise where a scalp electrode was used for the monitoring. Occasionally, accelerations or decelerations of the MHR were observed 1, 2. Recording of the MHR was more frequently noticed when the ultrasound transducer was used for fetal monitoring but the MHR pattern was not mentioned in the 49 cases encountered 3. Unintentional recording of the MHR was also described in four cases of delivery of the second twin but no decelerations of the MHR were observed 4. No change in the mean heart rate was observed in a meta-analysis on different maternal positions during cesarean section but changes in the MHR pattern were not investigated 5. However, van Veen et al. 6 reported on a study of 18 women where decreases in the MHR were observed during uterine contractions. These women were also in labor and received epidural anesthesia. Confirmed MHR and uterine activity were simultaneously recorded. A mean of 7.6 + 2.1 contractions per women were analyzed. MHR decreased by about 10 beats per minute (bpm) during contractions. In 56% (10/18) the decline in MHR resembled early decelerations of the fetal heart rate (FHR) 6.
Recently, another non-invasive technique for fetal monitoring was evaluated 7. The Monica AN24™ (Monica Health Care, Nottingham, UK) uses the fetal electrocardiogram, obtained from five high quality electrodes placed on the anterior abdominal wall of the mother. An advantage of the Monica AN24™ is that it also provides a continuous printout of the MHR and less MHR/FHR ambiguity when compared with the cardiotocograph 8. In addition, the Monica AN24™ can accurately record the FHR from as early as 20 weeks’ gestation 9. The Monica AN24™ also has the unique feature of using electrical impulses from the uterus to illustrate contraction patterns resembling what is observed with direct recordings of intrauterine pressure 10.
Stillbirths occur less frequently when mothers sleep on their left side 11 and fetal stress, as induced by venocaval compression during sleep, is a possible cause of unexplained late stillbirths 12. In addition, fetal oxygenation is lower in the supine position than in the left lateral position 13. Furthermore, the left lateral position is regarded as the most ideal position, as cardiac vagal activity is the least suppressed and sympathetic activity least enhanced 14.
The Safe Passage Study is a large prospective study to assess the effect of prenatal exposure to alcohol on stillbirths and sudden infant death syndrome 15. As part of the study, the FHR is monitored for at least 50 min under resting conditions with a 15° left or right lateral lift at 34–38 weeks’ gestation. In some of the recordings, regular decelerations of the MHR were observed, usually associated with uterine activity ( H.J. Odendaal, unpublished data). As MHR changes may be associated with the aortocaval syndrome, this study was done to determine the effect of different maternal positions on the MHR in late pregnancy before the onset of labor.
Material and methods
The study population comprised women booked for elective cesarean section at Tygerberg Hospital between February and June 2014. Women were recruited from antenatal wards on admission the day before the planned operation. Written informed consent was obtained as approved by the Health Research Ethics Committee of Stellenbosch University (approval number N13/10/153).
Relevant demographic information was obtained from the participant's medical record in addition to the body mass index and indication for the elective cesarean section. As far as possible, an early ultrasound examination was used to determine gestational age. If not available, the date of the last menstrual period was used. As a last alternative, the first symphysis pubis measurement was used.
Appropriately sized blood pressure cuffs were applied to the left upper arm and calf of the left leg. Four electrocardiogram electrodes of high quality were placed in a diamond shape on the anterior abdominal wall with the fifth electrode lateral to the one on the right side 7. Acceptable skin impedance for the electrodes was ensured. The participant lay in the supine position when the electrodes and blood pressure cuffs were applied but was turned on to her right or left lateral side before the recording was started. A computer program was used to randomize the side on which the recordings were started. Recordings of FHR, MHR and uterine activity were done continuously for 60 min, during which time the maternal position was changed every 15 min as illustrated in Figure 1. The mean of three blood pressure measurements was calculated as the mean for that period. If one blood pressure measurement was missed, the mean of two measurements was used as the value for that period. When the participant was lying on her back, the arm was resting at the side of her body. Lying on the left side, the arm was stretched out at a 45° angle from the rest of the body; when on the right side, the arm was placed straight, resting alongside the body. If any symptoms or signs of supine hypotension were encountered, the participant was immediately turned to the next lateral position.
After the recording, the Monica AN24 was connected to a dedicated personal computer where the DK 1.9 program was used to download and analyze the raw data for the MHR, FHR and uterine activity. Growth charts for Tygerberg Hospital were used to determine whether the newborn was small for gestational age 16.
Maternal heart rate, FHR, blood pressure recordings and uterine activity were analyzed separately for each of the four abovementioned 15-min periods. The Dawes–Redman criteria 17 of the Monica DK (version 1.9) software were used to analyze the raw data of the FHR for the mean heart rate, basal heart rate, small and large decelerations, small and large accelerations, and short-term variability according to standard definitions. Small and large accelerations were defined as having a duration of at least 15 s with amplitudes of 10 or 15 bpm, respectively. Small decelerations were defined as a deviation from baseline of at least 10 bpm for a minimum of 60 s and for large decelerations the duration had to be at least 30 s with an amplitude of 20 bpm or more. Due to the rapid changes in the MHR, it was not possible to determine the basal heart rate in some women. It was therefore not possible to differentiate between accelerations and decelerations. To select cases with obvious periodic changes it was therefore decided to define periodic changes as follows: four or more changes (from peak to nadir) in the MHR with an amplitude of at least 20 beats OR one change with an amplitude of 30 or more beats in at least one 15-min window of recording.
Statistical analyses were done with STATISTICA version 12 (StatSoft Inc., 2013; StatSoft Southern Africa - Analytics (Pty) Ltd, Sandton, South Africa). The distributions of variables were presented with categorized histograms and frequency tables. Maximum likelihood chi-squared statistics were used to analyze nominal data between categories. The relationships between continuous response variables (MHR) and nominal input variables (original position) over the different subsequent positions were analyzed using repeated measures analysis of variance (RMANOVA), with the compound symmetry assumption regarding the correlation-structure over time. Bonferroni multiple comparisons were used to compare the different treatment means over the recording positions at different times, i.e. “treatment × time” interactions.
Results
The study profile is given in Figure 1. Indications for the cesarean sections in these 119 mothers included previous cesarean section (n = 100, 84.0%), breech or transverse position of the fetus (n = 7, 5.9%), large baby (n = 4, 3.4%), placenta previa (n = 2, 1.7%) and other (n = 6, 5.0%). Many women had more than one indication for the cesarean section but one main indication was selected, usually a previous section. Demographic information is given in Table 1. There were 54 male and 65 female babies. In three mothers the cesarean section was done at 36 weeks’ gestation (birthweights 2570 g, 2950 g, 3190 g) and in one at 34 weeks (birthweight 2050 g). Four infants had birthweights of less than 2500 g: one born at 34, two at 37 and one at 39 weeks. Four newborns weighed less than the 10th centile on the weight for gestational age chart. Two were born at 37 weeks and the others at 39 and 40 weeks.
| Variable | n | Mean | Median | Minimum | Maximum | SD |
|---|---|---|---|---|---|---|
| Age (years) | 119 | 32.0 | 32 | 20 | 44 | 5.2 |
| Gravidity | 119 | 3.4 | 3 | 1 | 7 | 1.1 |
| Parity | 119 | 2.0 | 2 | 0 | 6 | 1.0 |
| Height (cm) | 119 | 159.4 | 158.6 | 141 | 180 | 6.8 |
| Weight (kg) | 119 | 87.4 | 85 | 48.3 | 150 | 22.2 |
| BMI (kg/m2) | 119 | 34.2 | 34.0 | 20.5 | 53.4 | 8.1 |
| GA (weeks) | 119 | 38.7 | 39 | 34 | 42 | 1.1 |
| Weight (g) | 115 | 3366 | 3340 | 1870 | 4710 | 539 |
| Apgar 1 | 115 | 8.4 | 9 | 3 | 10 | 1.2 |
| Apgar 5 | 115 | 9.4 | 9 | 8 | 10 | 0.6 |
| Apgar 10 | 114 | 9.8 | 10 | 8 | 10 | 0.5 |
- BMI, body mass index; GA, gestational age; SD, standard deviation.
Recordings started on the right side in 55 women and on the left side in 64, as randomly selected (randomization numbers were initially prepared for 240 participants). In two cases the recordings were discontinued before completion. In one case the woman did not want to participate any longer and in the other case the ward personnel had to start preparations for the cesarean section. However, the data that were recorded before the interruption were included in the analysis. In one participant the blood pressures in the leg could not be recorded because a large blood pressure cuff was not available. In another case, the blood pressure cuff was placed on the right leg instead because the left one had been amputated previously. In a third case the arm blood pressure was taken in the right arm because the left arm had a burn injury. In 26 cases the FHR could not be recorded during one period because of a signal loss. In six of these, the FHR pattern could not be extracted during any of the four periods.
When the recordings started on the right side, the lowest MHR (82.2 bpm) was observed in the left lateral position (Table 2). The decline in heart rate from the first supine position to the left lateral position and the increase to the second supine position were statistically significant (p < 0.0001).
| Maternal heart rate | Significance (Bonferroni test) | ||||
|---|---|---|---|---|---|
| Right lateral | Supine | Left lateral | Supine | ||
| n | 55 | 55 | 55 | 55 | |
| Mean heart rate (bpm) | 86.2a | 85.6b | 82.2a,b,c | 85.5c | ap < 0.0001, bp < 0.0001, cp < 0.0001 |
| SE | 1.5 | 1.5 | 1.4 | 1.4 | |
| −95% CI | 83.3 | 82.6 | 79.6 | 82.7 | |
| +95% CI | 89 | 88.5 | 84.9 | 88.4 | |
| Left lateral | Supine | Right lateral | Supine | ||
| n | 61 | 61 | 61 | 61 | |
| Mean heart rate (bpm) | 82.9a,c | 84.5b | 87.0a,b,c | 85.7c | ap < 0.001, bp < 0.0001, cp < 0.0001 |
| SE | 1.4 | 1.4 | 1.3 | 1.4 | |
| −95% CI | 80.2 | 81.8 | 84.4 | 82.9 | |
| +95% CI | 85.7 | 87.3 | 89.5 | 88.4 | |
| Systolic blood pressure arm | |||||
|---|---|---|---|---|---|
| Right lateral | Supine | Left lateral | Supine | ||
| n | 53 | 53 | 53 | 53 | |
| Mean SBP (mmHg) | 118.9a,b,c | 125.8a | 126.4b | 128.4c | ap < 0.0001, bp < 0.0001, cp < 0.0001 |
| SE | 1.9 | 2.0 | 2.2 | 2.0 | |
| −95% CI | 115.1 | 121.8 | 122.1 | 124.4 | |
| +95% CI | 122.7 | 129.8 | 130.6 | 132.4 | |
| Left lateral | Supine | Right lateral | Supine | ||
| n | 63 | 63 | 63 | 63 | |
| Mean SBP (mmHg) | 123.5a | 123.7b | 115.6a,b,c | 124.1c | ap < 0.0001, bp < 0.0001, cp < 0.0001 |
| SE | 1.8 | 1.8 | 2.0 | 1.8 | |
| −95% CI | 120.0 | 120.0 | 111.7 | 120.5 | |
| +95% CI | 127.0 | 127.3 | 119.5 | 127.7 | |
| Systolic blood pressure leg | |||||
|---|---|---|---|---|---|
| Right lateral | Supine | Left lateral | Supine | ||
| n | 54 | 54 | 54 | 54 | |
| Mean SBP (mmHg) | 141.8a | 145.2 | 144.3 | 147.5a | ap < 0.005 |
| SE | 2.2 | 2.4 | 2.4 | 2.3 | |
| −95% CI | 137.4 | 140.4 | 139.5 | 143.0 | |
| +95% CI | 146.3 | 149.9 | 149.2 | 152.0 | |
| Left lateral | Supine | Right lateral | Supine | ||
| n | 63 | 63 | 63 | 63 | |
| Mean SBP (mmHg) | 140.0 | 141.0a | 136.7a,b | 142.4b | ap < 0.05, bp < 0.001 |
| SE | 2.1 | 2.2 | 2.3 | 2.1 | |
| −95% CI | 135.9 | 136.6 | 132.2 | 138.3 | |
| +95% CI | 144.1 | 145.4 | 141.2 | 146.6 | |
- CI, confidence intervals; SBP, systolic blood pressure.
- a,b,cDifferences in p-values.
When the recording started on the left side, the lowest MHR (82.9 bpm) was also observed on that side (Table 2). The increase to and decrease from the right side were statistically significant (p < 0.0001). The heart rates were always the lowest when the women were lying on their left sides and were highest on their right sides.
The lowest systolic blood pressures were obtained when the women were lying on their right side (Table 2). The increases from the right side to the other three sides were statistically significant (p < 0.0001). Changes in the blood pressures in the left leg were less obvious. Lower values were again observed in the right positions (Table 2). When the recording started on the right side, the only significant difference was between this position and the last lateral position (Bonferroni p < 0.005). When the recording started on the left side, the only significant differences were from the supine to right position (Bonferroni p < 0.05) and from the right lateral to the last supine position (Bonferroni p < 0.001). The systolic blood pressure in the leg was lower than that in the arm during the supine position on one occasion in two participants, once during the first supine position and once during the second. Diastolic blood pressures in the arm and leg were also recorded and analyzed. As they followed the same changes as arterial pressures, the results are not given. Two women had symptoms of supine hypotension. One reported dizziness after assessments in all four positions had been completed. The other one reported nausea in the fourth period and was therefore turned onto her side. No hypotension was noted. There were no significant changes in the FHR, short term variability and the number of small accelerations between any of these positions.
Periodic changes in the MHR were observed in 13 women (10.9%). In 84.6% (11/13) of these women, the changes were associated with uterine activity. In 10 participants the changes were independent of the maternal position (Figure 2). The changes were limited to the lateral position in only one woman. In two the changes occurred in the supine position only (Figure 3). No significant differences were obtained between the two groups with and without periodic changes regarding systolic blood pressure measurements in the arm and leg, MHRs or FHR analyses (data not given). No significant differences regarding maternal demographic information, birthweight or Apgar scores were found between the two groups (Table 3).
| Variable | No MHR change, n = 106 | MHR periodic change, n = 13 | p | ||||
|---|---|---|---|---|---|---|---|
| n | Mean | SD | n | Mean | SD | ||
| Age (years) | 106 | 32.1 | 5.2 | 13 | 31.2 | 4.6 | 0.57 |
| Gravidity | 106 | 3.5 | 1.1 | 13 | 3.0 | 1.2 | 0.13 |
| Parity | 106 | 2.1 | 0.9 | 13 | 1.8 | 1.2 | 0.46 |
| Length (cm) | 106 | 159.4 | 7.0 | 13 | 159.7 | 5.7 | 0.89 |
| Weight (kg) | 106 | 88.6 | 22.5 | 13 | 77.8 | 18.5 | 0.10 |
| BMI (kg/m2) | 106 | 34.6 | 8.1 | 13 | 30.5 | 7.2 | 0.09 |
| GA (weeks) | 106 | 38.7 | 1.1 | 13 | 38.5 | 1.0 | 0.59 |
| Weight (g) | 102 | 3376 | 557 | 13 | 3290 | 372.8 | 0.59 |
| Apgar 1 | 102 | 8.4 | 1.2 | 13 | 8.0 | 1.5 | 0.26 |
| Apgar 5 | 102 | 9.4 | 0.6 | 13 | 9.1 | 0.8 | 0.06 |
| Apgar 10 | 101 | 9.8 | 0.5 | 13 | 9.5 | 0.5 | 0.06 |
- BMI, body mass index; GA, gestational age.
Discussion
Periodic changes in the MHR were only seen in 13 (10.9%) participants. Except for two mothers, the changes were always associated with uterine Braxton–Hicks contractions. There are two possible explanations for the observed changes in MHR with uterine contractions. During labor, contractions displace about 300–500 mL blood from the uterus and choriodecidual space into the venous circulation 18. This increase in the preload to the heart, according to the Frank–Starling mechanism, increases myocardial contractility 19. The heart rate may also increase but this is of short duration 18. During labor, additional stresses such as pain and bearing down efforts could also cause acceleration of the MHR. For obvious reasons, these accelerations are more common during the second phase of labor 6, 20. Information on the mechanism of accelerations of the MHR before labor is scanty. It is likely that the same mechanisms occur as during labor, but to a lesser degree. This is because less blood is auto-transfused during Braxton–Hicks contractions and they are not usually as painful.
The mechanism for periodic decline of the MHR is different, as aortocaval compression during the supine position could play a role 21. This compression could lead to decreased venous return to the heart and this subsequent decrease in preload could cause deceleration of the MHR 22, 23. It is most likely that venous return to the heart is reduced in a small proportion, as significant aortocaval compression only affects 2–4% of pregnant women 21. The most likely explanation for this is the existence of venous collateral circulations in the pelvis. The first is the important azygos system, a paired venous paravertebral pathway in the posterior thorax 24. The second is the collateral circulation along the ovarian veins, which expands greatly during pregnancy 25. If these alternative paths of venous return to the superior vena cava, upper inferior vena cava or left renal vein are not fully functional, venous return to the heart would be reduced during compression of the inferior vena cava. We postulate that reduced venous return heart could still occur in the lateral positions, as we also found decline in the MHR in these positions (Table 3, Figure 2). An explanation for this is that aortocaval compression probably occurs more often when the uterus contracts. When the uterus is relaxed it could mold around the great vessels, causing little compression, but during contraction, compression would happen more readily, even in the lateral position, as the firmer uterus would then compress the inferior cava by counter-pressure against the anterior abdominal wall.
The mean body mass index in mothers who had periodic decelerations of the MHR was indeed 4.1 kg/m2 lower, though not significantly lower (p = 0.09). This may imply that compression of the large vessel of the posterior abdominal wall could be more easily compressed by the firmer uterus in smaller pregnant women. It is also possible that pregnant women with smaller plasma volumes are more sensitive to changes in the pre- and afterload. However, this is more unlikely than decreased venous return as a possible mechanism for the periodic changes in MHR, as the large majority of newborns were appropriately grown and as 95% of cesarean sections were done for mechanical reasons. It is therefore unlikely that many mothers were more sensitive to an increase in plasma volume because of low plasma volumes.
We found that the pulse rate was 4 bpm higher when the participant was lying on her right side than on her left side. This is because vagal cardiac activity during late pregnancy is the least suppressed and the sympathetic activity is least enhanced on the left side. Aortocaval compression was thought to be the underlying mechanism 26. The fact that four cesarean sections were done before 37 weeks probably did not change the findings of the study. Birthweights of two of these infants were 2950 and 3190 g; it is therefore more likely that the gestational ages were wrong.
We found that the mean systolic blood pressure was the lowest in the upper arm. This was probably due to the effect of hydrostatic forces when non-invasive measurements are used in lateral positions. A previous study found that systolic blood pressure in the uppermost and dependent arms while in the lateral recumbent position are respectively 10 mmHg lower and 3.1 mmHg higher when compared with the supine position 27. The mean blood pressure in the leg was lower than in the arm in only two participants, indicating that aortic compression occurred rarely. We observed no decline in blood pressure in the supine positions but it is important to keep in mind that a reduction in cardiac output could occur in the absence of blood pressure changes 28, 29. We found no significant effect of position changes on the FHR and short-term variability.
The weak point in this study was that it was done in a selective population of normal pregnancies where the indications for the elective cesarean sections were predominantly mechanical. The study population was therefore not homogeneous. Pregnancies with placental insufficiency and lower blood volumes may therefore have been excluded. In addition, we did not study the direct association between cardiac output and periodic MHR changes. Furthermore, maternal positions were changed every 15 min, before the change in position could have had an effect on the FHR pattern.
The clinical relevance of the observed periodic changes in the MHR during Braxton–Hicks contractions in late pregnancy but before labor is still uncertain. We also could not differentiate between increases and decreases in the MHR because the basal heart rate was not always detectable. It is also uncertain whether periodic changes in the heart rate are caused by changes in the pre- or afterload of the heart secondary to uterine contractions affecting plasma volume.
Large epidemiologic studies are needed to determine the clinical significance of periodic changes in MHR during Braxton–Hicks contractions in late pregnancy.
Funding
Swedish International Development Cooperation Agency and Department of Obstetrics and Gynaecology, Stellenbosch University.
